Drake Bennett has a good piece about the idea of helping the poor by giving them money rather than through “aid programs” in which much of the appropriated funds end up being used (or, worse, misused) by aid providers and administrators rather than the intended recipients. The jury is hardly fully in on this subject, but the initial evidence seems reasonably promising.
My main concern about this would be that if you don’t design the program right, you’re going to in effect be instituting very high tax rates on the poor because they’re going to lose their eligibility for cash grants. Obviously that can be avoided by phasing the grants out very slowly or making them universal, but an idea that starts out looking like a pretty cheap way to help the poorest of the poor starts looking very expensive when you go big. For example, giving $150 a year to every citizen of Benin costs about $1.3 billion (admittedly, less than we give Israel). And of course at that point, cutting so many checks would probably just end up having weird distorting effects on the economy in terms of raising the price level rather than living standards.
That in turn is why it’s no coincidence that the big examples here at Brazil and Mexico — middle-income countries with a lot of inequality rather than very poor ones. And I think there’s a lot of evidence that this is the way for such countries to go. Rather than try to mitigate inequality through direct government intervention in the economy, try to adhere to the neoliberal orthodoxy and add on a generous dose of redistributive taxation.